How to repeat a block of lines using awk? - regex

Question

I'm trying to repeat the block of lines avobe the OCCURS word the number of times inticated in the line. The block of lines to repeat have a smaller number at the start of the line.
I mean, with this input:
01 PATIENT-TREATMENTS.
05 PATIENT-NAME PIC X(30).
05 PATIENT-SS-NUMBER PIC 9(9).
05 NUMBER-OF-TREATMENTS PIC 99 COMP-3.
05 TREATMENT-HISTORY OCCURS 2.
10 TREATMENT-DATE OCCURS 3.
15 TREATMENT-DAY PIC 99.
15 TREATMENT-MONTH PIC 99.
15 TREATMENT-YEAR PIC 9(4).
10 TREATING-PHYSICIAN PIC X(30).
10 TREATMENT-CODE PIC 99.
05 HELLO PIC X(9).
05 STACK OCCURS 2.
10 OVERFLOW PIC X(99).
This would be the output:
01 PATIENT-TREATMENTS.
05 PATIENT-NAME PIC X(30).
05 PATIENT-SS-NUMBER PIC 9(9).
05 NUMBER-OF-TREATMENTS PIC 99 COMP-3.
05 TREATMENT-HISTORY OCCURS 2.
10 TREATMENT-DATE OCCURS 3.
15 TREATMENT-DAY PIC 99.
15 TREATMENT-MONTH PIC 99.
15 TREATMENT-YEAR PIC 9(4).
15 TREATMENT-DAY PIC 99.
15 TREATMENT-MONTH PIC 99.
15 TREATMENT-YEAR PIC 9(4).
15 TREATMENT-DAY PIC 99.
15 TREATMENT-MONTH PIC 99.
15 TREATMENT-YEAR PIC 9(4).
10 TREATING-PHYSICIAN PIC X(30).
10 TREATMENT-CODE PIC 99.
15 TREATMENT-DAY PIC 99.
15 TREATMENT-MONTH PIC 99.
15 TREATMENT-YEAR PIC 9(4).
15 TREATMENT-DAY PIC 99.
15 TREATMENT-MONTH PIC 99.
15 TREATMENT-YEAR PIC 9(4).
15 TREATMENT-DAY PIC 99.
15 TREATMENT-MONTH PIC 99.
15 TREATMENT-YEAR PIC 9(4).
10 TREATING-PHYSICIAN PIC X(30).
10 TREATMENT-CODE PIC 99.
05 HELLO PIC X(9).
05 STACK OCCURS 2.
10 OVERFLOW PIC X(99).
10 OVERFLOW PIC X(99).
I tried it by this way:
tac input.txt |
awk '
BEGIN {
lbuff="";
n=0;
}{
if($0 ~ /^\s*$/) {next;}
if ($3 == "OCCURS") {
lev_oc=$1
len_oc=$4
lstart=0
for (x=1; x<n; x++) {
split(saved[x],saved_level," ")
if (saved_level[1] <= lev_oc) {
print saved[x]
lstart=x+1
}
}
for (i=1; i<=len_oc; i++) {
for (x=lstart; x<n; x++) {
print saved[x]
}
}
print $0
}else if ($0) {
saved[n]=$0
n++
}
}' | tac
But I don't get the result what I'm trying to obtain. Is awk the best way to do it? Do you have any alternative?

Answer 1

I used perl for this because it's easy to make arbitrarily complex data structures:
#!/usr/bin/perl
use strict;
use warnings;
# read the file into an array of lines.
open my $f, '<', shift;
my #lines = <$f>;
close $f;
my #occurring;
my #occurs;
# iterate over the lines of the file
for (my $i = 0; $i < #lines; $i++) {
# extract the "level", the first word of the line
my $level = (split ' ', $lines[$i])[0];
# if this line contains the OCCURS string,
# push some info onto a stack.
# This marks the start of something to be repeated
if ($lines[$i] =~ /OCCURS (\d+)/) {
push #occurring, [$1-1, $level, $i+1];
next;
}
# if this line is at the same level as the level of the start of the
# last seen item on the stack, mark the last line of the repeated text
if (#occurring and $level eq $occurring[-1][1]) {
push #occurs, [#{pop #occurring}, $i-1];
}
}
# If there's anything open on the stack, it ends at the last line
while (#occurring) {
push #occurs, [#{pop #occurring}, $#lines];
}
# handle all the lines to be repeated by appending them to the last
# line of the repetition
for (#occurs) {
my $repeated = "";
my ($count, undef, $start, $stop) = #$_;
$repeated .= join "", #lines[$start..$stop] for (1..$count);
$lines[$stop] .= $repeated;
}
print #lines;
For your reading pleasure, here's an awk translation.
BEGIN {
s = 0
f = 0
}
function stack2frame(lineno) {
f++
frame[f,"reps"] = stack[s,"reps"]
frame[f,"start"] = stack[s,"start"]
frame[f,"stop"] = lineno
s--
}
{
lines[NR] = $0
level = $1
}
# if this line contains the OCCURS string, push some info onto a stack.
# This marks the start of something to be repeated
$(NF-1) == "OCCURS" {
s++
stack[s,"reps"] = $NF-1
stack[s,"level"] = level
stack[s,"start"] = NR+1
next
}
# if this line is at the same level as the level of the start of the
# last seen item on the stack, mark the last line of the repeated text
level == stack[s,"level"] {
stack2frame(NR-1)
}
END {
# If there's anything open on the stack, it ends at the last line
while (s) {
stack2frame(NR)
}
# handle all the lines to be repeated by appending them to the last
# line of the repetition
for (i=1; i<=f; i++) {
repeated = ""
for (j=1; j <= frame[i,"reps"]; j++) {
for (k = frame[i,"start"]; k <= frame[i,"stop"]; k++) {
repeated = repeated ORS lines[k]
}
}
lines[frame[i,"stop"]] = lines[frame[i,"stop"]] repeated
}
for (i=1; i <= NR; i++)
print lines[i]
}

Answer 2

Here's a ruby solution:
#!/usr/bin/env ruby
# -*- coding: utf-8 -*-
stack = []
def unwind_frame(stack)
frame = stack.pop
_,occurs,data = *frame
with_each = stack==[] ? ->(l){ puts l} : ->(l){stack.last[2].push l}
occurs.times { data.each &with_each }
end
while gets
$_.chomp! "\n"
if m=$_.match(/OCCURS ([0-9]*)\.\s*$/)
puts $_
occurs=m[1].to_i
level = $_.to_i
stack.push([level,occurs,[]])
next
end
if stack==[]; puts $_; next; end
level = $_.to_i
if level > stack.last[0]
stack.last[2].push $_
next
end
while(stack!=[] && level <= stack.last[0])
unwind_frame(stack)
stack!=[] ? stack.last[2].push($_) : puts($_)
end
end
while(stack!=[])
unwind_frame(stack)
end
The result matches what you expected to get.